Data from: Colonisation dynamics during range expansion is poorly predicted by dispersal in the core range

The potential ranges of many species are shifting due to changing ecological conditions. Where populations become patchy towards the range edge, the realised distribution emerges from colonisation-persistence dynamics. Therefore, a greater understanding of the drivers of these processes, and the spatial scales over which they operate, presents an opportunity to improve predictions of species range expansion under environmental change. Species reintroductions offer an ideal opportunity to investigate the drivers and spatial scale of colonisation dynamics at the range edge. To this effect, we performed and monitored experimental translocations of water voles to quantify how colonisation and local persistence were influenced by habitat quality and occupancy. We used a novel statistical method to simultaneously consider effects across a range of spatial scales. Densely occupied neighbourhoods were highly persistent and frequently colonised. Persistence was more likely in high quality habitat, whereas influence of habitat quality on colonisation was less clear. Colonisation of suitable habitat in distant, sparsely occupied areas was much less frequent than expected from the well documented high dispersal ability of the species. Persistence of these distant populations was also low, which we attribute to the absence of a rescue effect in sparsely populated neighbourhoods. Our results illustrate a mismatch between the spatial scales of colonisation dynamics in the core and edge of a species range, suggesting that recolonisation dynamics in established populations may be a poor predictor of colonisation dynamics at the range edge. Such a mismatch leads to predictions of long lags between the emergence and colonisation of new habitat, with detrimental consequences for a species realised distribution, conservation status and contribution to ecosystem function. Conservation translocations that also reinforce existing populations at the range edge might stimulate the rescue effect and mitigate lags in expansion.

Latrine counts in 200 m waterway sections at each survey occasion
Latrine counts in 200 m waterway sections at each survey occasion: Data collected in the field. x = x coordinate (eastings) of section, y = y coordinate (northings) of section, xy = combined x and y coordinate (unique section ID), Date = dd/mm/yyyy, Count = count of latrines (that denote current occupancy) in section
Morgan_et_al_Data_Section_occupancy.csv

Habitat quality scores of 200 m waterway sections: Habitat quality score calculated from habitat variables collected in the field. x = x coordinate (eastings), y = y coordinate (northings), logitscore = habitat quality score
Morgan_et_al_Data_Section_habitat_scores.csv

Habitat variables for each 200 m waterway section: Data collected in the field. x = x coordinate (eastings), y = y coordinate (northings), width = waterway width (cm), ST_substra = stream bed substrate - 0 (silt, sand, gravel) or 1 (stones, boulders), pen1, pen2, pen3, pen4 = categorical variable for bank penetrability, tot_veg = percentage cover of vegetation, tree_cover = percentage cover by tree canopy, juncus = percentage cover by juncus sp., filip = percentage cover by Filipendual ulmaria
Morgan_et_al_Cell_coordinates_and_habitat_variables.csv

Colonisation status of waterway sections assuming 10 day closure period for analysis in R: Produced from raw data within R. occ.relax.F = colonisation status (1=colonised, 0=not colonised), ID.fac = waterway section ID, focal.x & focal.y = coordinates of section, logitscore = habitat quality score of focal section, disc.500 = mean habitat quality score of all sections within 500 m radius, disc.750 = mean habitat quality score of all sections within 750 m radius, yr.month = year-month combination, yr = year, month = month of year, Date.real.F = dd/mm/yyyy, meanmean.occ.N270 - meanmean.occ.N2500 = mean occupancy of all neighbouring sections within that distance band, hab.mean.N270 - hab.mean.N2500 = mean habitat quality score of all neighbouring sections within that distance band, conspec270to1000 = mean occupancy of all sections between 270 - 1000 m away
Morgan_et_al_Colonisation_Analysis_10_days_closure.csv

Colonisation status of waterway sections assuming 20 day closure period for analysis in R: Same as "Colonisation status of waterway sections assuming 10 day closure period for analysis in R", but variables derived assuming closure over 20 days
Morgan_et_al_Colonisation_Analysis_20_days_closure.csv

Colonisation status of waterway sections assuming 30 day closure period for analysis in R: Same as "Colonisation status of waterway sections assuming 10 day closure period for analysis in R", but variables derived assuming closure over 30 days
Morgan_et_al_Colonisation_Analysis_30_days_closure.csv

Persistence status of waterway sections assuming 10 day closure period for analysis in R: Same as colonisation analysis, except occ.relax.F = persistence status (1=remained occupied, 0=abandoned)
Morgan_et_al_Persistence_Analysis_10_days_closure.csv

Persistence status of waterway sections assuming 20 day closure period for analysis in R: Same as colonisation analysis, except occ.relax.F = persistence status (1=remained occupied, 0=abandoned)
Morgan_et_al_Persistence_Analysis_20_days_closure.csv

Persistence status of waterway sections assuming 30 day closure period for analysis in R: Same as colonisation analysis, except occ.relax.F = persistence status (1=remained occupied, 0=abandoned)
Morgan_et_al_Persistence_Analysis_30_days_closure.csv

This work is licensed under a CC0 1.0 Universal (CC0 1.0) Public Domain Dedication license.